AND01 troubles

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
User avatar
Lardo Boffin
Posts: 1221
Joined: Thu Aug 06, 2015 6:47 am
Contact:

Re: AND01 troubles

Post by Lardo Boffin » Wed Aug 29, 2018 9:00 pm

Just tried both drives with a model B with DFS with a working disc and basically both do the same.
The beeb works fine with my more modern drive.
Its a vanilla (other then being converted to UK spec) ANB09 so will have the older disc controller I believe. Not opened the case yet...
BBC model B 32k issue 4, 16k sideways RAM, Watford 12 ROM board, Viglen twin 40/80 5.25" discs, Acorn 6502 coproc
BBC model B 32k issue 7, turboMMC, Opus Challenger 3 512k, Pi 3 coproc
BBC Master, Datacentre + HDD, pi co-proc

Kazzie
Posts: 185
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: AND01 troubles

Post by Kazzie » Wed Sep 12, 2018 7:37 pm

As seen on the "I received..." thread, I have the two floppy drives, and this afternoon I've taken them to task.
IMG_20180912_1913595.jpg
Scrappy, Sticker, and No-Sticker
On the left is my drive, bought as faulty spares and repaired by me some months ago. Centre and right are LB's drives. I'll be referring to these (from left to right) as as Scrappy, Sticker, and No-Sticker. Their serial numbers (on the spindle motor) are 4343497, 4351852 and 4354061 respectively. Some differences between them: LB's ones have a head-load solenoid next to the spindle motor (which lifts or lowers the pressure pad that presses the disk to the read/write head) and two glass diodes instead of capacitors near the MC3470. Mine and the second of LB's contain U-S-C1 boards, whereas the third has a U-S-C1/2 board, with some further revisions to components and different routing for some of the soldered wires.
IMG_20180912_1914523.jpg
Solenoid hiding behind the motor. It's connected to the board by the two black wires seen on the near edge of the upper connector.

After checking that my scrappy was still working fine, I tested Sticker by reading a disk's catalogue. I didn't know about the solenoid, and the loud noise surprised me. I lifted the circuit board to get a good look at what was down there before going further. As well as the solenoid, I found this:
IMG_20180912_1335176.jpg
Debris under the pressure pad (replaced in order to take a photo)
What's that near the read/write head? Have a look:
IMG_20180912_1337445.jpg
Seems like Sticker is a good name for this drive.
We have a write-protect sticker, complete with backing paper. I wonder how long that's been in there? Anyway, back to the testing. Reading a catalogue, loading a BASIC formatting program, formatting the disk, and saving the program from memory to disk. "Sticker" did it all just fine. I even asked it to save eighteen copies of the short BASIC program to ensure it was able to step back and forth along tracks: no problem. Sticker appears to receive a clean bill of health, with the proviso that after I'd left the drive powered and Beeb running for three hours, it was a bit pernickety until I reset the Beeb. (Doing further testing with No-Sticker may have borked up some of the 8271's registers for all I know.)

Knowing that these drives hadn't been at all happy on the more modern Master, I changed the disk drive timing links on my Beeb to see how it behaved. My Beeb has an 8271 controller, and with both links in their default (open) configuration that gives a step time of 24ms, settle time of 20ms, and 64ms for head loading. (This is the setting specified for Olivetti 40 track drives in Acorn's DFS user guide.) I tried formatting a disk on the next slowest setting (6ms step, 50ms settle, 32ms head load) and the drive missed steps, didn't get back to track 0 to write the catalog, and generally sulked until I reset the power.

Different floppy controllers have different timing settings. Sprow's web page on floppy drives lists timings for the various link combinations for the 8271, 1770 and 1772, as well as the *CONFIG settings for the 1770 and 1772 on the Master. The 1770 settings include a reasonably slow option (30ms seek, 30ms settle) which should work for these old drives, but the slowest the 1772 will go is 12ms/15ms, which is certainly too fast for these drive to keep up with. Lardo, your US BBC B has an 8271 in it, but you may want to check the keyboard links, and perhaps check what chip is inside your Master too.

As for No-Sticker, this drive is a bit more troublesome. With the keyboard drive links set in dead-slow mode, it reads a disk catalogue and can load a file with no problem. If asked to format a disk, it will step all the way to track 39 and back to 0, but is fails to write the catalogue correctly. (It definitely writes something to the disk, as the disk is then unreadable in Sticker or Scrappy. Asking it to save a file to an already formatted disk is also problematic: it can sometimes overwrite a file okay, but asking it to save a new file leads to a corrupted catalogue again (and unreadable on the other drives).

So both drives' stepping motors are fine if given enough time, but No-Sticker seems to have some issue in its electronics somewhere. It doesn't appear to be a head alignment issue, as it can read other drives' disks and can't format its own. There's still more diagnosis to be done: I haven't started swapping circuit boards between the drives yet, but that's for another day.
Last edited by Kazzie on Wed Sep 12, 2018 7:37 pm, edited 1 time in total.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM (mid- restoration)
Acorn System 1 home-made replica

User avatar
Lardo Boffin
Posts: 1221
Joined: Thu Aug 06, 2015 6:47 am
Contact:

Re: AND01 troubles

Post by Lardo Boffin » Wed Sep 12, 2018 8:36 pm

Good stuff thanks!

Sticker was sold as working but it didn’t like any of my model Bs. I tried:-

Issue 4 - wouldn't switch on (suspected PSU not good enough)
Issue 7 - mostly worked but then wouldn’t load stuff off discs with loads in them
USA - as above

No-sticker was generally grumbly.
BBC model B 32k issue 4, 16k sideways RAM, Watford 12 ROM board, Viglen twin 40/80 5.25" discs, Acorn 6502 coproc
BBC model B 32k issue 7, turboMMC, Opus Challenger 3 512k, Pi 3 coproc
BBC Master, Datacentre + HDD, pi co-proc

Kazzie
Posts: 185
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: AND01 troubles

Post by Kazzie » Sun Sep 16, 2018 1:55 pm

A short update on further poking:

Swapping circuit boards between Sticker and No-sticker results in two grumpy drives. That's not too surprising, as Sticker has a revision 2 circuit board, whereas the others are revision 1.

Putting No-sticker's circuit board in Scrappy results in a floppy drive that passes all four tests: read catalogue, load format program, format floppy, save format program to disk. Doing the reverse, putting Scrappy's circuit board on No-sticker results in a floppy drive with the same write faults as No-sticker. Conclusion: No-sticker's fault isn't on the circuit board, but with the off-board components.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM (mid- restoration)
Acorn System 1 home-made replica

User avatar
Lardo Boffin
Posts: 1221
Joined: Thu Aug 06, 2015 6:47 am
Contact:

Re: AND01 troubles

Post by Lardo Boffin » Sun Sep 16, 2018 2:54 pm

Sounds like progress to me!
BBC model B 32k issue 4, 16k sideways RAM, Watford 12 ROM board, Viglen twin 40/80 5.25" discs, Acorn 6502 coproc
BBC model B 32k issue 7, turboMMC, Opus Challenger 3 512k, Pi 3 coproc
BBC Master, Datacentre + HDD, pi co-proc

Kazzie
Posts: 185
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: AND01 troubles

Post by Kazzie » Sun Sep 16, 2018 3:04 pm

Though with the occasional backward step: see the thread about the monitor my Beeb shares with my Archimedes :(
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM (mid- restoration)
Acorn System 1 home-made replica

Post Reply